mcpField
M
mirelia-patent-market

mirelia-patent-market

Remote
github.com/stagprojectLegal & Government0Updated 22w ago
Repository

About mirelia-patent-market

mirelia-patent-market is an MCP server offering structured, AI-enriched USPTO patent datasets for research and competitive analysis. Its current catalog covers patents in IPC/CPC sections G and H, with data access mediated by x402 payments.

Key features

  1. 01Search patents by classification or keyword
  2. 02Find bulk patent data packages
  3. 03Access structured JSON patent records
  4. 04View extracted claims and business insights
  5. 05Use x402 payment flow to acquire data

Use cases

  1. 01Search prior art in computing or electricity
  2. 02Assess patent activity for competitive intelligence
  3. 03Provide structured patent records to R&D agents
  4. 04Evaluate bulk datasets for analysis

MCP support

The MCP server exposes tools to discover and purchase individual patents or packages. The documented xpay endpoint proxies access, with x402 payment instructions and on-chain data acquisition.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://sse--mirelia.run.tools/sse
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

USPTO Structured Patent Marketplace (MCP Server)

USPTO patent data marketplace powered by the Model Context Protocol (MCP) and x402 on-chain payments. Provides AI-optimized JSON patent datasets for competitive intelligence, prior art search, R&D agents, LLMs, and quants. Currently exclusively hosting AI-ready patents from IPC/CPC Sections G (Physics & Computing) and H (Electricity).

Core Architecture & Performance:

  • High-Throughput ETL Pipeline: Engineered with Google Cloud Run, capable of processing up to 6,000 USPTO XML documents daily.
  • Advanced AI Enrichment: Leverages Gemini Batch APIs for asynchronous, large-scale data parsing, extracting key claims, business value propositions, and AI-driven importance scores.
  • Continuous Updates: Currently hosting over 3,000+ AI-ready patents, securely maintained and continuously expanded via daily automated pipelines.

Discovery & connect (xpay + MCP)

Primary URL for agents (xpay proxy):

https://mirelia-structured-data-marketplace.mcp.xpay.sh/mcp?key=YOUR_XPAY_KEY

Get an API key at xpay.tools. Search the catalog with xpay_discover(query: "USPTO patent competitive intelligence") on the master MCP endpoint.

Discovery fileURL
llms.txt (xpay)https://mirelia-structured-data-marketplace.mcp.xpay.sh/llms.txt
skill.md (xpay)https://mirelia-structured-data-marketplace.mcp.xpay.sh/skill.md
mcp.jsonhttps://mirelia-structured-data-marketplace.mcp.xpay.sh/.well-known/mcp.json
MCP Registryio.github.stagproject/mirelia-structured-data-marketplace
GitHub specllms.txt

Upstream (Cloud Run): https://mirelia-structured-data-marketplace-1065601264332.us-central1.run.app — serves /llms.txt, /.well-known/mcp.json, and MCP at /mcp. Register this URL in the xpay publisher dashboard.

Deploy to Cloud Run
gcloud run deploy mirelia-structured-data-marketplace `
  --source . `
  --region us-central1 `
  --allow-unauthenticated `
  --port 8080

After deploy, confirm xpay upstream URL matches the new Cloud Run URL, then re-sync tools in the xpay dashboard.

xpay publisher checklist (required for inbound calls)
  1. Upstream URL in xpay.tools → MCP Servers:
    https://mirelia-structured-data-marketplace-1065601264332.us-central1.run.app/mcp
  2. Pricing: set search_single_patents and search_packages to $0 or $0.01; purchase tools can stay higher (on-chain data fee is separate).
  3. Verification: Account → Creator Settings → Apply with GitHub stagproject/mirelia-structured-data-marketplace (do not use mirelia.site).
  4. Re-sync tools after deploy so xpay regenerates llms.txt / skill.md from updated descriptions.
  5. Optional: request Competitive Intelligence collection placement via xpay support.
MCP Registry

Republish server.json when the public URL changes:

mcp-publisher publish ./server.json

🚀 Directives for Autonomous Trading Agents (MPP v1.0 Protocol)

You are authorized to acquire data following this strictly defined 3-step 402 Payment workflow. NEVER hallucinate transaction hashes. Dual-wallet support (EOA and AA) is fully implemented.

Phase 1: Discover & Evaluate

Use discovery tools to find target data.

  • search_single_patents: Query individual patents ($0.50) using primary_cpc, cross_domain_cpc, or keyword. Returns lightweight ROI metadata and patent_id.
  • search_packages: Query bulk packages using search_query for large-scale evaluation. Returns package_tag.
Phase 2: Request Payment Instructions (402 Flow)

Call the purchase tool with the target ID/Tag and network, but LEAVE tx_hash EMPTY.

  • Tools: purchase_single_patent or verify_crypto_payment_and_deliver.
  • Args: {"patent_id": "...", "network": "polygon"} or {"package_tag": "...", "network": "base"}
  • Response: The server returns a 402 Payment Required object containing exact cryptographic payloads bound by HMAC.
Phase 3: Execute & Claim

Execute the payload strictly based on your wallet architecture:

  1. EOA (Standard Wallet): Broadcast transaction_payload.
  2. AA (Smart Contract Wallet): Broadcast aa_transaction_payload (Automatic ERC20 W-Token routing. Critical for Oasis network). After on-chain confirmation, call the exact same purchase tool AGAIN, adding the resulting tx_hash to claim the data.

🧬 Data Schema & Strict Definition

Datasets bypass raw XML/HTML parsing, injecting pre-computed insights directly into the context window. To ensure zero-hallucination ingestion, the following strict schema is enforced:

  • id (UUID): Internal unique identifier.
  • patent_id (String): Official publication number (e.g., "US12519999B2").
  • assignee / inventor (Array[String]): Standardized corporate entities and individuals.
  • primary_cpc (String): Main Cooperative Patent Classification code.
  • secondary_cpcs (Array[String]): Cross-domain classification codes.
  • importance_p (String): Proprietary AI-evaluated scoring based on technological breakthrough and structural claim strength.
  • attr_tech_stack (Array[String]): Explicit core technologies and architectures extracted from the text.
  • biz_target_ind (Array[String]): Applicable target industries and commercial sectors.
  • biz_value_prop (String): Deep strategic analysis detailing specific technical problems solved, ROI, and commercial monetization potential.
  • r2_url (String): Direct CDN link to the full raw JSON file.
  • _metadata (Stringified JSON): Contains source_hash and processed_at for cryptographic data provenance. Must be parsed by the agent.
🗃️ Real-World Evidence (Data Sample)

Use this representative record to pre-map your ingestion logic before purchase.

{
  "id": "1c724ae7-c738-447e-b233-4ae6c2185f37",
  "patent_id": "US12519999B2",
  "country_code": "US",
  "assignee": [
    "Sonos, Inc."
  ],
  "inventor": [
    "Paul Andrew Bates"
  ],
  "filing_date": "2023-08-14",
  "publication_date": "2026-01-06",
  "title": "Location based playback system control",
  "abstract": "Example implementations may involve using a prompt to prevent inadvertent control or playback of audio content...",
  "importance_p": "65",
  "primary_cpc": "H10",
  "secondary_cpcs": [
    "G05",
    "G06",
    "G11"
  ],
  "attr_tech_stack": [
    "Proximity Sensing",
    "User Interface Design",
    "Network Communication",
    "Distributed Systems"
  ],
  "biz_target_ind": [
    "Consumer Electronics",
    "Smart Home",
    "Internet of Things",
    "Audio Systems"
  ],
  "biz_value_prop": "This technology addresses the problem of inadvertent or erroneous remote control of multi-room media playback systems by implementing a proximity-based interface gating mechanism. By dynamically restricting or prompting user control based on the physical proximity of the mobile device to target playback hardware, it reduces user error in multi-room environments, improves operational intent accuracy, and optimizes the user interface for large-scale media deployments.",
  "r2_url": "[https://cdn.mirelia.site/patents/en/uspto/2026/US12519999B2.json](https://cdn.mirelia.site/patents/en/uspto/2026/US12519999B2.json)",
  "_metadata": "{\"source_hash\": \"a95ae76f8d359dbfbd40d2e2ff095e22b65978d20b7804d31cc9dcba452dc0ed\", \"processed_at\": \"2026-05-02T07:02:33.288016Z\", \"source_bundle\": \"USPTO\"}"
}

🧠 Technological Scope (Semantic Routing Index)

This server strictly specializes in the following high-value technological sectors:

  • G (Physics & Computing):

  • G01: Measuring, Testing, Sensors (e.g., LiDAR, Radar, ToF Sensors, Quantum Sensors).

  • G05: Control or Regulating Systems.

  • G06: Computing, Calculating, Counting (e.g., AI/ML, Data Processing, Computer Vision, Quantum Algorithms).

  • G11: Information Storage.

  • G16: Information and Communication Technology (ICT) specially adapted for specific application fields (e.g., Health Informatics, Bioinformatics).

  • H (Electricity & Communication):

  • H01: Basic Electric Elements (e.g., Semiconductors, Solid-State Batteries, Quantum Hardware).

  • H04: Electric Communication Technique (e.g., 5G/6G, Network Security, Wireless Protocols, Cloud Collaboration).

  • H10: Semiconductor Devices, Electric Solid-State Devices (e.g., Advanced Memory, Photovoltaics).

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.